Combined Actions of Chemicals in Mixture doc

10 Combined Actions of Chemicals in Mixture 10.1 INTRODUCTION Following the current practice, health risk assessments of exposure to chemicals and the subsequent regulatory measures, e.g., classification and labeling, establishment of regulatory standards such as Maximum Residue Limits (MRLs), etc. are generally based upon data from studies on the individual substances. However, humans are simultaneously exposed to a large number of chemicals that potentially possess a number of similar or different toxic effects. Consequently, not only opponents against the use of chemicals but also the consumers at large are increasingly challenging the authorities to consider that this ‘‘chemical cocktail’’ or ‘‘total chemical load’’ does not produce unforeseen health effects. This question was even more highlighted in 1996 when the U.S. Congress passed the U.S. Food Quality Protection Act (FQPA) (US-EPA 1996). This act requires that the US-EPA consider the effects of exposure to all pesticides and other chemicals that act by a common mechanism of toxicity when tolerances for pesticide use in crops are derived. Therefore, the aspect of combined actions of chemicals needs to be addressed to a greater extent in the risk assessment process. A major obstacle in doing so is the lack of data from studies on chemical mixtures employing generally accepted toxicological methods, such as short- and long-term animal studies. This is because most of the limited resources in experimental toxicology are used to study single chemicals or the effects of pretreatment with one chemical on the effects of another. In addition, data on human exposures to chemical mixtures are in general very inadequate. Therefore, regulatory agencies are faced with the situation that they cannot always reliably predict whether the simultaneous exposure to foreign chemicals in the environment and food constitutes a real health problem. As the possible combin- ations of chemicals are innumerable and experimental testing of all such mixtures is not feasible for obvious reasons, there is a need for science-based advice on how exposure to mixtures of chemicals can be dealt with in the risk assessment. Interactions between chemicals administered to humans at high doses have been known for many years in the field of pharmacology. However, these experiences are not directly useful for predicting toxic effect s of mixtures of environmental chemicals because the exposure levels for the general human population are relatively low and interactions occurr ing at high doses may not be representative for low-dose exposures (Könemann and Pieters 1996). Toxicity studies with mixtures have bee n performed for several decades. Initially, most studies were done with binary mixtures. Later, studies with defined mixtures of more than two compounds have been reported. Studies have also been performed with complex mixtures of environmental chemicals, such as exhaust condensates, in order to gain insight into the toxic effects of such a particular mixture. However, the interpretation of the toxicity seen in these latter studies is complicated because the exact composition of the mixtures is normally not known, and the ‘‘real life’’ mixtures may v ary considerably in composition. Therefore, extrapolation to other situations may be difficult. This fact is often ignored for the sake of simplicity. ß 2007 by Taylor & Francis Group, LLC. A major issue in the asses sment of the combi ned toxi cological effect of chemicals in a mixture is the type of combi ned action to be expect ed. What kind of toxicity may be expect ed, given the toxi city pro fi les of the indi vidual co mponents? Bliss (1939) was the first to provide a conc eptual fram ework for the combi ned acti on of chemicals and later contr ibutions wer e made by Finney (1942) , Hew lett and Placket (1959, 1964), Placket and Hewle tt (1952, 1963, 1967), Ash ford and Cob by (1974) , and Ashford (1981) . Placket and Hewle tt (1952) provi ded a scheme of possibil ities of combi ned (joi nt) action s, see Tab le 10.1. A major clue that can be taken from this schem e is that , in the initial assessmen t, it is imp ortant to evalua te whet her inte ractions are actual ly occurr ing (present) o r not (absent). Inter action was de fi ned a s the infl uence of one chemi cal on the biological acti on of anothe r, eith er quali tatively or quantitat ively. The schem e represents the extre mes of combined actions. In many cases, adequat e informat ion about the underl ying mechan isms of combi ned actions is not avail able. This led Berenbaum (1985, 1989) to propose three class es of combi ned acti on: zero inte raction, synerg ism, and antago nism. Cur rent know ledge about the combined toxi cological effects that may occur from exposur es to diff erent ch emicals in mixtures is outlined in this chapter. Sp ecial atte ntion is paid to the low levels of exposur es norm ally encounte red from the unint ended, indi rect exposure to chemical mix tures throu gh food and environmen t. It should be recognized that it has not been possible to cover all possi ble combi ned exp osures to chemi cals in this book. The mai n empha sis is paid to the ident i fication of the basic princ iple s for combined actions and inte ractions of chemi cals (Secti on 10.2), and to the curren t know ledge on effects of expo sures to mix tures of industri al chemi cals, incl uding pesticide s and environmen tal contam inants. Test stra te- gies to assess c ombined acti ons and interacti ons of chemicals in mix tures (Secti on 10.3) as well as toxi cological test met hods (S ection 10.4) are addres sed, approac hes used in the asses smen t of chemi cal mixtures are presen ted (Sectio n 10.5), and examples of ex periment al studies using simple, well-d e fined mixtures are given (Section 10.6). 10.2 BASIC CONCEPTS AND TERMINOLOGY USED TO DESCRIBE THE COMBINED ACTION OF CHEMICALS IN MIXTURES The major objective in the risk assessment of exposure to mixtures of chemicals is to establish or predict how the resulting toxicological effect might turn out. Will the toxic effect be determined by simple additivity of dose or effect, or will it deviate from additivity, either by an effect stronger or less than expected on the basis of additivity? The prediction of the toxicological properties of a chemical mixture requires detailed informa- tion on the composition of the mixture and the mechanism of action of each of the individual compounds. In order to perform a risk assessment, proper exposure data are also needed. Most often TABLE 10.1 Classification of Combined (Joint) Toxic Actions of Two Compounds in Mixture Combined Action Interaction Similar Action Dissimilar Action Absent (No interaction) Simple similar action (Dose addition) Independent action (Response addition) Present (Interaction) Complex similar action (Antagonism or synergism) Dependent action or complex dissimilar action (Antagonism or synergism) Source: Modified from Placket, R.L. and Hewlett, P.S., J. Royal Statistical Society, Series B 14, 143, 1952. ß 2007 by Taylor & Francis Group, LLC. such detailed infor mation is not avail able. Com plex chemi cal mixtures may contai n hundred s, or even thous ands o f compo unds, and thei r compo sition is quali tatively and quantitativel y not full y know n and may change over time. Adeq uate testing o f such mix tures is most often imp ossible because they are either virtuall y unavailabl e for test ing or only avail able in such a limited amoun t that a suf ficient n umber of dose levels cannot be applied. In ad dition, high- dose levels of a chemical mixture may have diff erent types of effects than low dose level s an d high- to low-dos e extra polation may be meani ngless. In the foll owing, severa l terms used to descri be interactions between chemi cals are mentioned as well as basic c oncepts used in the haza rd and risk asses sment of chemi cal mix tures. The descripti on of these basic concepts, first o utlined by Bliss (1939) and Placket and Hewlett (1952), are based on the publicati ons by Köne mann and Pieters (1996) , Cassee et al. (1998) , and Groten et al. (2001) . The de finitions of addit ivity, synerg ism, antago nism, and potentiat ion are those of Klaassen (1995) and Seed et al. (1995) . As has already bee n outlined in the introduct ion, one of the main point s to consi der is whether there will be no inte raction or interacti on in the form of either synerg ism or antago nism. These three basic principle s of combined acti ons of chemical mixtures are purely theor etica l and one often has to deal with two or all three concepts at the same time, especi ally when mix tures consist of more than two compo unds an d when the toxi city targets are more complex. Inter actions between ch emicals may be of a physico-chem ical a nd=or biolog ical natur e. Exampl es of physico-chem ical inte ractions are the reaction of nitrite wi th alkyl amines to produce carci nogenic nitr osamines, and the bindin g of toxic chemi cals to active charcoal resulting in a decreas ed absorp tion from the gastr ointesti nal tract . It is held that physi co-chem ical inte ractions will norm ally o nly occur at high doses and therefore are of less er importan ce for low-dos e scenar ios. Physic o-chemical interactions will there fore no t be considered in a ny detai l in this book. 10.2.1 NO INTERACTION Acco rding to Placket and Hewle tt, there are two types of combin ed action without interacti on (Table 10.1): sim ple sim ilar action (dose addition, Loe we additivi ty) and sim ple diss imilar acti on. This latter type contains two conce pts: effect or respon se additivity and Bliss indepe ndence. The indepe ndence criterio n seem s not to be wide ly used in toxic ology (Groten et al. 2001). The respon se to a mixture of compo unds depends not only on the dose, but also on the correlati on of tole rances between the effect s of the chemi cals in the mix ture, which can vary between À1 and þ 1 (Bliss 19 37). There is a compl ete negative correlati on ( r ¼À1) betw een the effects of two ch emicals if the individua ls that are most suscep tible to one toxicant are least suscep tible to the other , while a complete posit ive correlatio n ( r ¼þ1) exists if the individua ls most suscep tible to one toxicant are also most suscep tible to the other . 10.2.1. 1 Simpl e Simi lar Action (Dos e Addi tion, Loe we Additiv ity) Simp le sim ilar action (simple joint action or co ncentrati on=dose addit ion) is a nonint eractive proces s in whi ch the chemicals in the mix ture do not affect the toxicit y of one anothe r. All the chemi cals of concern in the mixture act on the same biological site, by the same mechanism of action, and differ only in their potencies. The correlation of tolerances is completely positive (r ¼þ1) and each chemical contributes to the toxicity of the mixture in proportion to its dose, expressed as the percentage of the dose of that chemical alone that would be required to obtain the given effect of the mixture. Thus, the individual components of the mixture act as if they were dilutions of the same toxic compound and their relative potencies are assumed to be constant throughout all dose levels. An important implication is that, in principle, no threshold exists for dose additivity. Simp le sim ilar action serves as the basis for the use of toxi c equiva lency facto rs (T EF, Secti on 10.5.1.4) often used to descri be the c ombined toxic ity of isomers or structura l analog ues. Additi ve ß 2007 by Taylor & Francis Group, LLC. effects are described mathematically using summation of doses of the individual compounds in a mixture adjusted for differences in potencies. This method is assumed to be only valid for compounds that produce linear dose–response curves. Probably, the best validated example of a group of compounds that obey the principles of simple similar actions are the dioxins (polychlorin- ated dibenzo-p-dioxins and dibenzofurans) that produce most (if not all) of their toxicities through interaction with the Ah-receptor. 10.2.1.2 Simple Dissimilar Action (Response or Effect Additivity, Bliss Independence) Simple dissimilar action (simple independent action, independent joint action, Bliss independence, and effect addition or response addition) is also a noninteractive process where the toxic effect of each chemical in the mixture is not affected by the other chemicals present. However, the modes of action of the constituents in the mixture will always differ and possibly, but not necessarily, the nature and site of action also differs among the constituents. Response addition is referred to when each individual of a population (e.g., a group of experimental animals or humans) has a certain tolerance to each of the chemicals in a mixture and will only exhibit a response to a toxicant if the concentration exceeds the tolerance dose. In such a case, the number of responders within the group will be recorded rather than the average effect of a mixture on a group of individuals. By definition, response addition is determined by summing the responses of the animals to each toxic chemical in the mixture. Three different concepts have been developed for effect=response additivity depending on the correlation of susceptibility of individuals to the toxic agents: . Complete Negative Correlation There is a complete negative correlation between the effects of two chemicals if the indivi- duals that are most susceptible to one toxicant are least susceptible to the other. This is the simplest form of response additivity. The proportion (P) of individuals responding to the mixture is equal to the sum of the responses to each of the components: P mixture A,B ¼ P A þ P B less than or equal to 1 . Complete Positive Correlation There is a complete positive correlation between the effects of two chemicals if the individuals most susceptible to one toxicant are also most susceptible to the other. The proportion (P)of individuals responding to the mixture is equal to the response to the most toxic compound in the mixture: P mixture A,B ¼ P A if toxicity A ! B . No Correlation This situation is equal to Bliss independence. There is no correlation if the proportion of individuals responding to the mixture is equal to the sum of proportions of indiv iduals responding to each of the toxicants taking into account that those individuals that respond to constituent A cannot react to B as well: P mixture A,B ¼ P A þ P B Á (1 À P A ) Although this type of correlation seems to be similar to complete negative correlation, the difference is that, in this case, an individual can respond to both compounds A and B but not to both at the same time. ß 2007 by Taylor & Francis Group, LLC. The approac h of respon se addit ion can be easily applied to simple p roblems, such as acute toxicit y of pesticide s. However , more complex effects are not always easy to summ ate. Experi- mental animals are usually obtained from inbre d stra ins, while human popula tions are more heter ogeneous . In a ddition, various effect s on d ifferent o rgan syst ems may occur withi n different time frames in experi mental animals. The US-EPA (1986) appli ed the concept of response addit ion to the deter minati on of cancer risks, assuming a compl ete negative correl ation of tolerance. This assum ption is consi dered to contr ibute to a conser vative estimat ion of risk , since the correlati on of tolerances may not be stric tly negative in inbred homog enous experi mental animals. The re is a major difference between the concept s of respon se addition and dose addit ion when the human situatio n of low exposur e levels is asses sed. Response addition imp lies that when d oses of che micals are below the no-eff ect levels of the indi vidual compo unds (i.e., the respon se of each chemical equals zero), the combi ned acti on of all compo unds together will also be zero. In co ntrast, dose add ition can a lso occur below the no-eff ect level and the combi ned toxi city of a mixture of compo unds at individua l levels b elow the no-eff ect level may lead to a respon se. For compo unds with presumed linear dose –respon se curves , such as genoto xic and carci nogeni c compo unds for which it is assumed that a no-eff ect level does not exist and for which the mecha nism of a ction may be regarded as simil ar, respon se addition and dose addition wi ll provi de identical results (Könemann and Pieters 1996). 10.2.2 INTERACTIONS:COMPLEX SIMILAR ACTION AND COMPLEX DISSIMILAR ACTION Chemicals in mixtures may interact with one another and modify the magnitude and sometimes also the natur e o f the toxi c effect . As illust rated in Tab le 10.1, the combi ned acti on of chemicals that interacts can be divided into two categories: complex similar action and complex dissimilar action (dependent action). Interactions may take place in the toxicokinetic phase and=or in the toxicodynamic phase. The interactions may result in either a weaker (antagonistic) or stronger (potentiated, synergistic) combined effect than would be expected from knowledge about the toxicity and mode of action of each individual compound. . Antagonism An antagonistic effect occurs when the combined effect of two chemicals is less than the sum of each chemical given alone. Synonyms sometimes used for antagonism are interaction, depotentiation, desensitization, infra-addition, negative synergy, less than addi- tive, subaddition, inhibition, antergism, competitive antagonism, noncompetitive antagonism, uncompetitive antagonism, or acompetitive antagonism. . Synergism A synergistic effect occurs when the combined effect of two chemicals is greater than the sum of the effects of each chemical given alone. Synonyms sometimes used to describe synergism are: coalitivity, interaction, uni-synergism, augmentation, sensitization, supra-addition, inde- pendent synergism, dependent synergism, degrada tive synergism, greater than additive, co-synergism, super- addition, conditional independence, or potentiation. . Potentiation Potentiation, being a form of synergism, occurs when the toxicity of a chemical on a certain tissue or organ system is enhanced when given together with another chemical that does not have toxic effects on the same tissue or organ system. This form of interaction is especially well described in mutagenesis and carcinogenesis where a number of compounds have been identified as co-mutagens or co-carc inogens. ß 2007 by Taylor & Francis Group, LLC. The ultimate toxicological response following exposure to a chemical substance is most commonly the result of the action of this substance on a definite site or receptor. For a given concentration of the agent at the target site, the intensity of the response will depend on the quality of the action (the intrinsic activity) and the affinity of the compound for the receptor. When two compounds exert the same action by acting at different sites, their interaction will often result in a synergistic effect but a simple additive effect is also a possibility (the synergism between smoking and asbestos exposure is the classical example). 10.2.2.1 Complex Similar Action In the case of complex similar action, two compounds acting on the same target receptor do not produce an additive effect as would be expected from simplicity, but either an antagonistic or synergistic effect. This phenomenon is well known for substances competing for the same hormonal or enzy matic receptor sites. In such cases, lower than additive effects are often observed. An example could be two chemicals that exert the same action (e.g., accumulation of acetylcholine) by acting in the same manner (e.g., by inhibition of acetylcholine esterase). An additive effect may occur if the intrinsic activities and affinities of the two substances are identical but most often an antagonistic effect is observed as both compounds compete for the same receptor. A maximal antagonism is found when the substance with the lowest intrinsic activity possesses the higher affinity for the receptor or has been the first to get into contact with the target. In order to predict the effect of a mixture of chemicals with the same target receptor, but with different nonlinear dose–effect relationships, either physiological or mathematical modeling can be applied. For interactions between chemicals and a target receptor or enzyme, the Michaelis– Menten kinetics (first order kinetics but with saturation) are often applicable. This kind of action can then be considered a special case of similar combined action (dose addition). It is highly likely that, for compounds thought to have complex similar actions, the observed deviations from the expected additivity in some cases are due to the fact that the compounds are actually not acting at the very same site at the target receptor. This means that the compounds actually have complex dissimilar actions and the combined action is misclassified as a complex similar action due to insufficient knowledge about the exact mechanisms of action. 10.2.2.2 Complex Dissimilar Actions Complex dissim ilar actions are probably the most frequently occurring interactions operating in experimental studies on mixtures applying high doses. The most obvious cases in the toxicokinetic phase involve enzyme induction or inhibition. Enzyme induction could result in a synergistic effect if more reactive (and toxic) intermediates are formed or in an antagonistic effect if the toxic agent is removed by detoxification. Compoun ds which influence the amount of biotransformation enzymes can have paramount effect on the toxicity of other chemicals. Uptake and excretion are often active processes which may also be affected by other chemicals. Interaction between substrates for the same membrane receptors or pumps, as well as for biotransformation enzymes could result in synergism and antagonism, too. 10.3 TEST STRATEGIES TO ASSESS COMBINED ACTIONS AND INTERACTIONS OF CHEMICALS IN MIXTURES Ideally, all chemicals in a mixture should be identified and the toxicity profile of each of the constituents as well as their potential combined actions and=or interactions should be determined over a wide range of exposure levels. For complex environmental mixtures, this approach is not realistic and therefore a number of approaches and test scenarios have been presented to obtain toxicological information on mixtures with a limited number of test groups (Cassee et al. 1998). ß 2007 by Taylor & Francis Group, LLC. 10.3.1 TESTING OF WHOLE MIXTURES Althoug h testing of the whole mix ture as such seems to be the proper way to approac h the risk asses sment of exposure to that mix ture, it wi ll not provi de data on c ombined actions and=or interacti ons betwee n the indi vidual compo nents of the mixture. Eve n if the effect of the mixture is compa red with the effects of ea ch individua l compo nent at compa rable con centratio ns, this will not a llow a descri ption of p otential synerg ism, potent iation, or antago nism, and it is even d oubtful that deviations from additivi ty can be concluded. This can o nly be achiev ed if dose –respon se curves are obtained for each of the singl e compo unds. Testi ng of the whole mix ture as such has been recommend ed for mix tures that are not well charact erized (Mum taz et al. 1993), and h as succes sful ly been applied for asses sing the combi ned toxicit y of simple, de fin ed chemi cal mixtures wher e the toxi cological proper ties of the individua l compo nents wer e also investig ated, see Sectio n 10.6. 10.3.2 PHYSIOLOGICALLY B ASED T OXICOKINETIC MODELING For many chemi cals, their metaboli sm is the maj or deter mina nt of the risk and for a number of hazardo us compoun ds, there is a consi derabl e kno wledge from experiment al studi es on the rela- tionship between metaboli sm and toxicit y. In particula r, in vitro studies using cell cult ures, subcel lular fractions , or pure enzymes have provided infor mation on the nature of reactive inter- mediates as well as on d etoxi ficati on pathways . Moreov er, the signi ficance of these proces ses has been demon strated in severa l speci es of experi mental animals and human s. Physio logicall y Based Tox icokineti c (PBTK) model s are deriv ed similarly to Physio logicall y Based Pharmacok inetic (PBP K) model s, which h ave been used for a numbe r of years in the develo pment of medi cinal drugs. They describe the rat or man as a set of tis sue compa rtments , i.e., liver, adipos e tissues, poorl y perfus ed tissues, and richl y perfused tissues along with a descripti on of metaboli sm in the live r. In case of v olatile organi c compo unds a descrip tion o f gas exchange at the level of the lung is included, see also Secti on 4.3.6. In principle, the in vivo human metabolism can be predicted by using in vitro enzyme kinetic data and can thus be compared with the in vitro and in vivo data from experimental animals. For example, experiments using microsomes or hepatocytes may predict the in vivo velocity of metabolism for a single metabolic pathway. Such data may be incorporated in PBTK modeling (Andersen et al. 1995, Leung and Paustenbach 1995, Yang et al. 1995). As a rule, the description of the rate constants such as V max and K m for the individual (iso)enzymes follows Michaelis–Menten kinetics. Therefore, interindi- vidual differences in expression levels of enzymes and genetic polymorphism can also be modeled. Ploemen et al. (1997) have presented a strategy to combine PBPK modeling with human in vitro metabolic data to explore the relative and overall contribution of critical metabolic pathways in man. In order to use PBTK modeling in the assessment of mixtures, Cassee et al. (1998) suggest that one of the components is first modeled and regarded as the prime toxicant being modified by the other components. Based on in vitro data on the other components, effects of, e.g., inhibition or induction of specific biotransformation isoenzymes can be incorporated in the model. Effects of competition between chemicals in a mixture for the same biotransformation enzymes may also be incorporated by translating the effects into effects on the Michaelis–Menten parameters that are then incorporated into the model. PBTK models can potentially be extended to include the toxicodynamic phase (PBTK=TD model) if a direct relationship exists between the concentration of the active metabolite (or parent compound) and the toxic effect (Yang et al. 1995). 10.3.3 ISOBOLE METHODS An isobole is a contour line representing equi-effective quantities of two agents or their mixtures (Loewe and Muischnek 1926). ß 2007 by Taylor & Francis Group, LLC. The theoretic al line of a dditivity is the straight line that connects the individua l doses of each of the singl e agents that produce a predet ermined, fi xed effect alone, for example a n ED 50 (50% respon se) o f a given toxi c or bioche mical effect. The isobo le method is widely used to evalua te the effects of binar y mix tures. However , a large numbe r of different mixtures of the two compo unds have to be tested in order to identify combi nations that produce the fixed effect . If the graphi cal repres entation (iso bologram) of the combi nations that produce the fixed effect show s a stra ight line, the two compo unds behave in a dose-a ddit ive manne r and subseq uently, can be regarded as compo unds that have a similar mode o f action, see Figure 10.1. In c ase of an antago nistic inte raction all the e qui-effect concent rations in the mixtures repres ent an upwa rd concave line in the isobologr am (Figur e 10.2), wher eas a synerg istic interacti on woul d produce a down ward curve in the isobologr am (Figur e 10.3). In practice, the interpretation of test results strongly depends on the accuracy of the estimated intercepts of the theoretical isobole with the axis, which represents the doses of the single compounds that induce the desired effect. In fact, large standard deviations of these intercepts prevent a reliable conclusion as to the deviation from additivity. Berenbaum (1981) introduced an equation to calculate an interaction index (CI). This enables the effects of noninteractive combinations to be calculated directly from dose–effect relationships of the individual compounds, regardless of the particular types of dose–effect relations involved. CI ¼ d 1 =D 1 þ d 2 =D 2 þ : þ Ld n =D n d 1 , d 2 , , d n are the doses of the agents in the mixture D 1 , D 2 , , D n are the doses of the individual agents producing the same effect as the mixture Dose agent A (mg/kg bw) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Dose agent B (mg/kg bw) 1 2 3 4 5 6 7 8 9 Isobole ED 50 of agent A is 5 mg/kg bw ED 50 of agent B is 10 mg/kg bw Experiments showed that 4 mg/kg bw A + 2 mg/kg bw B produce ED 50 3 mg/kg bw A + 4 mg/kg bw B produce ED 50 2 mg/kg bw A + 6 mg/kg bw B produce ED 50 1 mg/kg bw A + 8 mg/kg bw B produce ED 50 FIGURE 10.1 Isobologram of two agents A and B that act additively. ß 2007 by Taylor & Francis Group, LLC. Dose agent A (mg/kg bw) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Dose agent B (mg/kg bw) 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 Isobole ED 50 of agent A is 5 mg/kg bw ED 50 of agent B is 10 mg/kg bw Experiments showed that 4 mg/kg bw A + 6.30 mg/kg bw B produce ED 50 3 mg/kg bw A + 8.65 mg/kg bw B produce ED 50 2 mg/kg bw A + 9.50 mg/kg bw B produce ED 50 1 mg/kg bw A + 9.82 mg/kg bw B produce ED 50 FIGURE 10.2 Isobologram of two agents that act antagonistically. Dose agent A (mg/kg bw) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Dose agent B (mg/kg bw) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Isobole ED 50 of agent A is 5 mg/kg bw ED 50 of agent B is 10 mg/kg bw Experiments showed that 4 mg/kg bw A + 0.18 mg/kg bw B produce ED 50 3 mg/kg bw A + 0.50 mg/kg bw B produce ED 50 2 mg/kg bw A + 1.35 mg/kg bw B produce ED 50 1 mg/kg bw A + 3.70 mg/kg bw B produce ED 50 FIGURE 10.3 Isobologram of two agents A and B that act synergistically. ß 2007 by Taylor & Francis Group, LLC. For binary mixtures, a straight line (isobole) is produced joining D 1 and D 2 and passing through (d 1 , d 2 ). The interaction index (CI) is 1, <1, or >1 when the combinations show zero interaction, synergism, or antagonism using dose addition, respectively. In cases of departure from additivity, the magnitude of CI depends on the ratio of the concentrations of the constituents of the mixture. Thus the CI is not a general figure but depends on the specific concentrations of the chemicals in the mixture. One difficulty in using this approach is to determine when a specific CI actually deviates from 1 (additivity), as the method of isoboles as developed does not include measures to decide whether deviations from the line of additivity are systematic or simply due to chance or experiment al error (Cassee et al. 1998). One way of dealing with this problem is to calculate con fidence intervals for the iso-effective doses of the single compounds and to add a confidence belt to the line of additivity (Kortenkamp and Altenburger 1998). This envelope of additivity is an area in which those combinations of two compounds are lying that has a specific effect and may reasonably be considered as showing no interaction (for details see Cassee et al. 1998). The isobole method can also be applied to mixtures where only one of the two agents produces the effect under consideration. In case agent A produces an effect, whereas agent B does not, the equation is reduced to CI ¼ d 1 =D 1 ¼ 1 In this case the iso-effective dose, D 2, of the agent lacking the effect of interest can be regarded as infinitely large , so that the resulting additivity isobole runs parallel to the respec tive dose axis. Combination of three agents can be analyzed by constructing three-dimensional isobolar surfaces, and combinations of more than three compounds can be assessed more easily by using a generalization of the above-mentioned equation. However, new procedures using a polynomial model have been proposed to evaluate more complex mixtures (Cassee et al. 1998). One of the strengths of the isobole method is that it can be used to analyze combined effects of compounds irrespective of the shape of their dose–response curves. It is possible to assess mixtures of agents with dissimilar dose–response relationships, even when the maximal effects are not identical (Kortenkamp and Altenburger 1998). Although isoboles are very illustrative, a compl ete construction requires a large amount of data sets both on the single compounds and mixtures and large standard deviations may limit the interpretation ( Cassee et al . 1998) . However , even if it is desirable to test combinations of agents at several m ixture ratios so that the iso boles can be constructed reliably, Kortenkamp and Altenburger (1998) are of the opinion that this is not always a necessary prerequisite. Valid conclusions about the combination effect of mixtures can often be drawn on the basis of surprisingly few data. 10.3.4 COMPARISON OF INDIVIDUAL DOSE–RESPONSE CURVES Comparison of dose–response curves of one chemical (A) in the absence and presence of a second chemical (B) has been proposed as a tool to predict whether the combined action of the two chemicals is either additive or independent (Cassee et al. 1998). In the case of dose additivity, the dose–response curve of A is determined on a linear- or log-dose scale, and an equi-effective dose of A (d A,equi ) and B (d B ) resulting in the same effect is estimated. Using the fixed dose d B of chemical B and adding various doses (d A – d A,equi ) of A, the dose–response curve should shift to the left and reach the same maximum as the maximum for the dose–response curve of A alone when the effect of B is smaller than A max . However, in case of competitive agonism, the effect of B does not affect the effect of A þ B at higher dose of A. ß 2007 by Taylor & Francis Group, LLC. [...]... thereby the toxicity of components in the mixture The tissue doses of chemicals in mixture can be predicted with PBTK models when the binary interactions between all of the components in the mixture are known (Haddad et al 1999a,b, 2000a,b) However, the quantitative characteristics of each of these binary interactions have to be determined by experimentation Given the complexity of the mixtures, to which... toxicity of parathion and paraoxon by acting as a pool of noncritical enzymes, which compete for the binding of paraoxon thereby preventing an inhibition of cholinesterase The increase in the level of carboxyl esterase and cholinesterase has the potential to enhance further the ability of toxaphene to limit the toxicity of parathion The authors therefore anticipated the toxicity of a mixture of parathion... for mixtures which consist of a large number of widely varying chemicals with no obvious ranking of individual constituents according to their potential health risks and the top-ten chemicals of the mixture are not easily identified This approach involves identification of the top-ten classes of chemicals to be lumped together by class to the top-ten chemicals to be treated as a simple mixture The lumping... Food Research Institute in Zeist, the Netherlands, initiated a research program to obtain some basic information on the toxicological interactions between toxicologically well-characterized chemicals in well-defined mixtures The objective was to establish knowledge about some general principles for the interaction of chemicals in mixtures that would be useful in the risk assessment of complex mixtures The... administration in drinking water Oral administration in diet In the main study, a number of effects were observed at the LOAEL and, based on the toxicity data of the individual compounds, most of these effects were expected A few effects seen in the toxicity studies with the individual compounds had disappeared in the combination, whereas some effects not seen in the range-finding studies with the individual... whereas the one obtained using Vmax and Km values was in between The kinetic data from mixture exposures were within the simulated boundaries of blood concentrations However, with increasing complexity of the mixtures, the impact on the blood kinetics of the single components became progressively more important, i.e., blood concentrations of unchanged parent chemicals increased with mixture complexity... methods available to study combined actions and toxicological and biochemical interactions of chemicals in mixtures are essentially the same as those used for the study of single chemicals in order to examine their potential general toxicity and special effects such as mutagenicity, carcinogenicity, and reproductive toxicity One especially successful method of testing complex mixtures is bioassay-directed... than that of parathion Thus the results of the study could indicate an antagonistic effect of toxaphene on parathion and on paraoxon Chaturvedi (1993) also examined the effect of mixtures of 10 pesticides (alachlor, aldrin, atrazine, 2,4-D, DDT, dieldrin, endosulfan, lindane, parathion, and toxaphene) administered by oral intubations or by drinking water on the xenobiotic-metabolizing enzymes in male... the toxicity of the mixture without identifying the type of interactions between the individual components However, the results of the testing can only be used for hazard characterization following exposure to that particular mixture A more detailed approach is to assess the combined action of the individual components in the mixture Several experimental designs can be used, primarily depending on the... Weight -of- Evidence Modification to the Hazard Index The HI method does not incorporate information on interactions among components of the mixture (ATSDR 2004) Mumtaz and Durkin (1992) proposed a weight -of- evidence (WOE) method to systematically address this need The method was designed to modify the HI to account for interactions, using the weight of evidence for interactions among pairs of mixture . study combined actions and toxicological and biochemical interactions of chemicals in mixtures are essentially the same as those used for the study of single chemicals in order to examine their. enzymes could result in synergism and antagonism, too. 10.3 TEST STRATEGIES TO ASSESS COMBINED ACTIONS AND INTERACTIONS OF CHEMICALS IN MIXTURES Ideally, all chemicals in a mixture should be identified. occurring interactions operating in experimental studies on mixtures applying high doses. The most obvious cases in the toxicokinetic phase involve enzyme induction or inhibition. Enzyme induction